V O L U M E 2 0 , N O . 1, J A N U A R Y 1 9 4 8 Table XII.
9
Paraffin and Cycloparaffin Hydrocarbon Content of Catalytic Cracked Naphtha Boiling between 28' and 120" C. 1
Hydrocarbon
28-49
2,Z-Dimethb lhutane 2,3-D1methvlbutane 2-Nethylpentane 3-hfethylpentane n-Heuane 2,2,3-Trimethylhutane 2-Methylhexane 3-Wethylhexane n-Heptane 2,2,4-Trimethylpentsne 2,5-Dimethylhexane 2,4-Dimethylhexane 2,2,3-Trimethylpentane 2,3,4-Trimethylpentane 2,3-Dimethylhexane 4-Methylheptane 1 3,4-Dimethylhexane 2-Methylheptane 3-Methyl-3-ethylpentane 3-Ethylhexane Cyclopentane Methylcyclopentane Cyclohexane 1,l-Dimethylcyclopentane 1 3-Dimethylcyclopentane 1:2-Dimethyloyclopentane} Methyleyclohexane Ethylcyclopentane Trimethylcyclopentane Other Ca cyclopareffins
1
0.14 0.72
..
.. ..
.. .. .. .. .. .. ..
..
Fraction . Xi0 . 5 6 7 Boiling Point, ' C. 49-67 67-82 82-93 93-99 99-103 103-111 2
3
.. o:ii
0.14 4.57
18.07
10.52 1 . 1 3 3:81 . . . 1.03
.. ..
.. ..
. . . . . .....
4
. . .
. . . . . . . . . .. o:ii .. 5.81
6.02
..
..
.. .. . . . . . . . . . . . . . . . . . . . . . . . . . . .
0:47 0.18
.. .. .. .... .. ..
. . . . .
. . . .
. . . . . . . . . . . . .
.
...
.. . . ... ...
Total
LITERATURE CITED
28-120
.. .. .. .. ..
(1) d m . Petroleum I n s t . , A.P.I. Project 6 Report, "Analysis of Alkylates and H y d r o codimeis" (August 31, 1946). ( 2 ) Bond, G. R., Jr., ISD. ESG. CHEM.,XSAL.ED.,18,692 (1946). (3) Brewer, A . K., and Dibeler,
..
...
*.
.. ..
..
.... .. ..
...
..
.. ..
0 : 47 0.58 0.26
,.
0.10
.. .. .. .. ..
..
.. .. ..
.. .. .. .. ..
.. .. ..
....
... ... ...
...
... ... ...
... .. ... 0.81 1.06
..
.. .. ..
0.36 1.47 1.04 1.31 1.37
. .
... ...
..
.. .. .. .. .. ..
0.52 0 . 8 0 9:48 1124 0.67 0.06 .. 0 . 5 6 0.71 6.75 1.90 0.79 0.28 2.58 0.10 0.80 0.06
... ... .. .. .. ..
9 118-120
..
1:53 1 . 4 3 0 : 94 0.20 0.05
..
..
8 111-118
Volume Per Cent
. . . .
unavailability of pure calibrating samples.
,. .. .. .. .... ..
0.28 5.29 18.18 10.52 4.94 1.14 5.81 7.55 2.37 0.26 0.47 0.58 0.26
.. ..
0.91 1.06
i:E]
3.06
4.68
... ... ... ...
.. .. ..
0.99 11.70 0.73 1.27 9.80 4.33 2.32 1.37 3.20
... 0 : 38
...
1.83
.. .. .. ..
.. ..
LOO. 00
V. H., J . Research Natl. BUT. Standards, 35, 125 (1916). (4)
Mair, E. J., a n d Forziati, A.
F., Ihid., 32, 165 (1944). (5) Taylor, R. C . , and Young, IT.'. S.,IXD. EXG. CHEM., ANLL. ED., 17, 811 (1945). (6) Washburn, H. W.,Wiley, H . F., and Rock, S. M., Ihitl., 15,541 (1943). (7) Washburn, H. IT.'., Wiley, H. F., Rock, S. M., and Berry, C. E., Ibid., 17, 74
(1945).
c6and
heavier olefins and cyclo-olefins cannotbe directly on the mass spectrometer because of pattern similarity and
RECEIVEDMarch 28, 1947. Presented before the Division of Petroleum Chemistry a t the 111th Meeting of the AMERICAKCHEVICAL SOCIETY, Atlantic city, N. J,
Infrared Analytical Techniques for Analyzing C5 Mixtures VERNON THORNTON AND ANNETTE E. HERALD, Phillips Petroleum Company, Bartlesrille, Okla. Mixtures of Cj hydrocarbons have been analyzed in the vapor phase rapidly and with sufficient precision for plant control. The analytical techniques were greatly simplified by the discovery that the extinction per unit pressure was constant over a pressure range of 50 to 650 mm. of mercury at 30" C. for the t w o Cj
W
HEN the problem of applying infrared analytical techniques to the control of plant operations was considered for streams containing mixtures of Cb hydrocarbons, the litcrature had little to offer in the way of developed methods. The difficulty of confinhg these volatile samples in conventional liquid absorption cells is evident and most laboratories have spent considerable effort toward building a satisfactory liquid cell capable of containing samples under small pressures (1-4). I n view of this difficulty it was decided to investigate the possibility of analyzing Cg mixtures in the vapor phase. To this end highly purified samples of each hydrocarbon listed in Table I R ere scanned a t desirable pressures in conventional type gas absorption cells of suitable length over the rock salt range of a Perkin-Elmer Model 12-A spectrometer equipped with photoamplifier and Brown recorder. From these scannings, extinctions were measured a t t u entythree wave lengths, corresponding to key absorption bands of the components. These extinctions, measured at several pressures, were plotted against uncorrected pressures and in every case a straight line through the origin resulted. Figure 1 shows this linear relationship for a paraffin, n-pentane a t the 8.7-micron
paraffins and six Cj olefins under consideration. Small variations in the concentrations of components appearing in fractionation products were quickly measured by infrared method, even though the components possessed relatively weak absorption bands, by comparing with a reference mixture.
band, for an olefin, 2-methyl-1-butene at the 8.2-micron band, and for a mixture of 2 - m e t h y l - l - b ~ t e n e ( 5 2 ~and ~ ) 1-pentene (48y0), a t the 8.2-micron band. This linear relationship n'as the most that could be hoped for and somewhat surprising to one who had analyzed C4mixtures by a similar procedure. Key wave lengths were chosen from the spectrograph records and pressure-extinction curves plotted for each of the eight materials shoL5-n in Table I a t each of the key TTave-lengt,h positions. From the slope of these curves the extinction coefficients needed to set up t,he usual set of simultaneous equations were obtained. Table I1 s h o w one such set of extinction coefficients. The underscored extinction coefficients are those of the principal absorber at the spectral positions indicated. The most unfavorable Table I. Compound 3-Methyl-1-butene Isopentane 1-Pentene 2-lfethyl-1-butene
Components of a Cj Cut
Boiling Point, O C. 18.8 27 89 30.1 31.05
Compound trans-2-Pentene n-Pentane cis-2-Pentene 2-Methyl-2-butene
Boiling Point, O C. 35.85 36.0 37.0 38.49
ANALYTICAL CHEMISTRY
10
component appearing in amounts of less than 1%. Obviously, special methods must be em2ployed to handle such require\Taw 3XethylLength, MethylMi1Isu12-Methyltrans-2cis-22ments. Components crons butene C1 Pentene 1-butene n-Cs Pentene Pentene butene As one example, in a frac0.92 0.72 3-Methyl-1-butene 1 4 . 6 33.8 0.36 3.51 1.04 101.0 13.7 tionator producing a mixture 8.5 8.87 1o.j 4.90 2.02 3.41 4.47 1.82 3.12 Iaopentane 2.86 6.37 with the composition shown in 1,33 3.35 1,57 1-Pentene 5.45 16.8 1.72 18.0 2-Methyl-1-butene 11.4 5.27 0.82 36.6 136.0 5.24 7.M 4.52 5.00 Table IV, it was necessary to n-Pentane 13.75 1.39 0.32 4.16 1.32 16.3 1.66 41.32 1.70 control the tower so that the trans-2-Pentene 10.36 20.0 7.14 22.5 5.67 1.07 1610.. 05 5 l 9 . E6 4 18.0 n-pentane component remained cia-2-Pentene 14.4 43.0 0.44 3.32 1.12 1.15 1.38 less than 0.5%. 1.11 5.89 10.93 81.0 2-Methyl-2-butene 1 2 . 5 8.12 3.58 2.43 6.76 A conventional five-component infrared analysis was neither fast enough nor accurate enough to be used for this control. The tower was controlled, however, by comparing a sample drawn off the liquid phase of an intermediate tray every 2 hours with a sample having the approximate composition of the tray but with zero n-pentane content. By measuring the extinction at the spectral position of the key absorption band of n-pentane first for the standard sample and then for the tray sample without altering the instrument settings, small changes in the n-pentane content of the tray sample could be detected. Since it had previously been determined that the n-pentane content of the sample should be kept between 0 and 2'%, the column engineers could keep the column upcrating properly from results handed them 20 minutes after sampling. The results of 7 days' successful operation are shoa n 'E----in Figure 2, in which the per cent of n-pentane determined a t 100 200 300 400 50D 600 approvimately 2-hour intervals, as described above, is plotted PRESSURE IN MM HG against time. Figure 1. Pressure-Extinction Curves Illustrating Table 11.
Extinction Coefficients of CI Hydrocarbons ( E / P Lx 10 -9
+
Linearity A.
Gpara5n.
B.
C' Olefin.
C.
Crmixture
case is that of 3-methyl-1-butene in the presence of cis-2-pentene. Fortunately, these two compounds differ in boiling point by 18" C. and there is little chance of finding both in the same fraction. Comparison of results obtained on different days with different operators on the same sample are shown in Table 111. The time involved in sampling, instrument running, and computing was about 1.5 hours. Sample pressures and path lengths were chosen to give extinction measurements between 0.250 and 0.650 whenever possible. The range of workable pressures n.as 50 to 650 nim. of mercury a t 30" to 33 ' C. Cell lengths used were 5 and 10 cm. To control fractionating columns adequately, the conccntration of a key component a t some point in the toxer may be desired at frequent intervals and in such cases the above method of analysis beconies too slow. To this requirement may be added the request for a more accurate determination of a
Table 111. Components Isopentane 1-Pentene 2-Methyl-1-butene n-Pentane trans-2-Pentene ris-2-Pentene 2-1Iethy-l-2-butene
Table IT.
Comparison of Infrared Analyses (Lab. 4.UT.R.No. 55) Analyzed 8-29-46 17.4 54.2 1.4 11.6 3.7 2.6 5.5
.4nalyzed 8-30-46 18.4 53.7 1.6 12.4 3.7 2.6 5.7
Coniposition of Product from Fractionator %